CN101427531B

CN101427531B - A protection method of the speakers of the interdomain protocol in the optical network

Info

Publication number: CN101427531B
Application number: CN2007800143304A
Authority: CN
Inventors: 孙德胜; 王哲
Original assignee: ZTE Corp
Current assignee: SHANDONG CHENGDA NEW ENERGY TECHNOLOGY Co.,Ltd.
Priority date: 2007-01-04
Filing date: 2007-01-04
Publication date: 2011-11-23
Anticipated expiration: 2027-01-04
Also published as: WO2008086641A1; CN101427531A

Abstract

There is a protection method used in the speakers of the interdomain protocol in the optical netwrok. In the step S102, a master speaker and a backup speaker in the domain are chosen, each of which gathers the node information and routing information in the domain and performs inter-layer communication. In the step S104, the master speaker establishes communication with the backup speaker, and performs routing information critical data backup and status backup. In the step S106, the master speaker and the backup speaker check the living status of each other. If the backup speaker detects the fault of the master speaker, it becomes into the new master speaker and a new backup speaker is chosen, then return to the step S104. If the master speaker detects the fault of the backup speaker, a new backup speaker is chosen, then return to the step S104.; If fault doesn't appear in both speakers, then repeat the step S106. The speakers can be efficiently protected, rapidly switched. And the algorithm is less complicated, compact and reliable.

Description

The optical-fiber network inter-domain protocol is represented guard method a little

Technical field

What the present invention relates to optical-fiber network represents some backup and protection, relates in particular to a kind of optical-fiber network inter-domain protocol that is used for and represents a little guard method.

Background technology

Optical-fiber network, OTN (Optical transmission network for example, optical transport network), WDM (Wavelength-division multiplexing, wavelength division multiplexing), SDH (Synchronous digitalhierarchy, synchronous digital hierarchy) or SONET (Synchronous optical network, Synchronous Optical Network) transmits net, be used widely at field of telecommunications.

ASON (Automatic switched optical network is called for short ASON) is meant the optical-fiber network of new generation of finishing automatic function of exchange under routing and signaling control, is the research focus of field of optical networks in recent years.The introducing of ASON technology can strengthen the quick allocative abilities of Network, improves professional survivability, effectively resists network multiple spot fault, and the different grades of service can be provided flexibly, satisfies the needs of the differentiated service that develops rapidly at present.ITU-T G.8080 and G.771X series suggestion has proposed the notion of ASON and has realized framework.

Along with the progressively deployment of ASON network, different operators, distinct device supplier's network will be interconnected by external network interface (external network network interface is called for short E-NNI).Interconnected these networks of E-NNI all will be as routed domain independently.From application point of view, the connection in the optical-fiber network is relatively-stationary, and does not need frequent foundation and deletion on the one hand, therefore not necessarily all needs routing iinformation, particularly inter-domain routing information on each apparatus for network node.In addition, for the consideration of management and safety, operator generally can use the exchange that corresponding strategies is controlled routing iinformation between the territory.

But for same operator or same equipment supplier's network, though that all devices all passes through internal network interface (internal network network interface, be called for short I-NNI) is interconnected.In same territory, ASON network using ospf (open shortest path first is called for short OSPF) agreement is carried out the node route maintenance.But consider the complexity of efficient, routing management and maintenance, often need to be divided into several sub-routed domain.Like this for a plurality of routed domain, how collaborative work is finished route from the whole network angle and is calculated and safeguard, seems extremely important.G.7715, ITU-T has then stipulated ASON network inter-domain routing protocol (Domain toDomain Routing Protocol with OIF2003.259.00/2002.023.10, be called for short DDRP) framework, by with the routed domain layering with abstract route information is made a summary and gathered, introducing is represented point (being called for short SPEAKER) and has been solved this problem, helps the expansion of network.

SPEAKER converges the nodal information and the routing iinformation of place routed domain, and the SPEAKER of notice last layer or following one deck.So, the node that the SPEAKER of different layers just can obtain the whole network can reach information, has solved the problem that OSPF can not be cross-domain.

The status of SPEAKER is extremely important.If have only a SPEAKER in the territory, in case this SPEAKER breaks down, then the routing inquiry function and the information interaction of interlayer are interrupted fully, the DDRP agreement need be re-elected SPEAKER, and a series of attended operations such as carry out that relevant information floods, the time is long, and efficient is low.At this moment, finishing the connection recovery by Routing Protocol if desired can cause the business recovery time to prolong greatly.

Therefore, need a kind of technical scheme, can carry out perfect SPEAKER protection.

Summary of the invention

Main purpose of the present invention is to provide a kind of optical-fiber network inter-domain protocol that is used for to represent a little guard method, is used for solving the protection problem of existing DDRP agreement SPEAKER.

To achieve these goals, the invention provides a kind of optical-fiber network inter-domain protocol that is used for and represent a little guard method.This guard method may further comprise the steps:

Step S102, main with representing a little and backup is represented a little in the election territory, the master uses to represent a little and back up and represents nodal information and the routing iinformation that a little converges the territory, place respectively, and the execution interlayer communication;

Step S104, main with representing a little and backup establishes a communications link between representing a little, be used to carry out that the routing iinformation critical data backs up and state backs up;

Step S106, main with represent a little and back up represent a little between by communicating to connect mutual detection survival condition separately, a little occur fault if backup is represented a little to detect to lead with representing, then backup is represented a little becomes new master with representing a little, elects new backup to represent a little, be back to step S104, if the master fault a little occurs with representing a little to detect to back up to represent, then elect new backup to represent a little, be back to step S104, do not break down if all detect, then repeated execution of steps S106.

In step S102, election is main with representing a little and backup is represented a little and can be may further comprise the steps:

Step a, the candidate qualification of disposing each node, include qualification become main with represent point, qualified become to back up represent point or do not have qualification;

Step b, between node and node, set up after two-way communication is connected, each node is listed whole main using that may become respectively and is represented a little and back up the tabulation of representing node a little, each node is according to the candidate qualification then, declare oneself to be main represent point or do not make a speech with representing point or backing up, if it oneself is main with representing a little or backing up and represent a little not having the node statement, then change step e over to;

Step c, whole qualified becoming in the main node listing of using in representing a little, creating one does not also have declaration to become the main subclass of representing a little of using;

Steps d oneself is main with representing a little if one or more node is all stated, the node that has limit priority so becomes the master with representing a little, determines that by comparing other attributes final master with representing a little, proceeds to step f if priority equates;

Step e, the node that selection has limit priority is main with representing a little, relatively other attributes are determined finally main with representing a little if priority equates;

Step f, whole qualified becoming in the node listing of backup in representing a little, creating one does not also have to declare to become and backs up the subclass of representing a little;

Step g oneself is that backup is represented a little if one or more node is all stated, the node that has limit priority so becomes backup to be represented a little, determines that by other attributes relatively final backup represents a little if priority equates, flow process finishes;

Step h, selection has the node of limit priority to be represented a little for backup, relatively the definite final backup of other attributes is represented a little if priority equates.

Above-mentioned other attributes can comprise node identification and IP address.

In step S106, the survival condition that main usefulness detects separately between representing a little and back up and representing a little mutually can may further comprise the steps:

Step P1 main starts a first timer with representing, and is used to follow the tracks of detect the time interval of sending STATUS ENQUIRY message, and backup is represented and started a second timer, and whether be used to detect has inquiry or order to send with representing a little from the master in the stipulated time;

Step P2 if first timer at first then, then enters step P3, if second timer at first then, then enters step P6;

Step P3 main a little send out message and represents a little to backup with representing, and inquires about and backs up the state of representing a little, and start the 3rd timer, is used to inquire about whether to back up the response time of representing a little overtime, and the time-out first timer;

Step P4 a little receives the main Query Information of representing a little of using if backup is represented, and then backup is represented and a little sent response, and the second timer that resets has simultaneously shown that inside receives query message at the appointed time;

Step P5, if the 3rd timer then before, the master backs up the response message of representing a little with representing a little to receive, then the master a little closes the 3rd timer with representing, being illustrated in the stipulated time receives response message, if and message is represented backup to represent dotted state normal, then main with representing a replacement first timer, show that backup represent point all are normal, be back to step P1, if also do not receive then that at the 3rd timer response or the response that receives represent that backup represents a mistake, then the master does not a little recover first timer with representing, judge the master and fault a little occurs with representing, suspend and detect, execution is switched;

Step P6, backup representing that the inquiry that a little initiates a message is main to suspend second timer with the state of representing a little, and starts the 4th timer and come to use the response of representing a little to monitor to the master;

Step P7, if the 4th timer then before, backup is represented and is a little received the main response message of representing a little of using, and message shows that main all are normal with representing point, and then backup is represented and a little closed the 4th timer, recovers and restarts second timer, be back to step P1, fault a little occurs otherwise judge to back up to represent, suspend and detect, execution is switched.

In step S106, backup is represented a little becomes new master with representing a little, elects new backup to represent a little and can may further comprise the steps:

Step T1, backup is represented and is a little at first checked the main with the database of representing a little of local backup;

Step T2, if the database of backup in representing a little is effective, then backup is represented a little to send to the node that exists and is stated that oneself is the message of representing a little, and waits for the answer of other nodes, if invalid, then proceeds to step S102;

Step T3 receives after the answer of other network nodes, and the state of a change itself is represented in backup, and the state of oneself is set to new master with representing a little;

Step T4, the node of other in network sends updating message to new master with representing a little, waits for database update;

Step T5, new master represent a little to send to back up and represent a node election message, back up the election of representing a little again.

Elect new backup to represent a little and can may further comprise the steps:

Whole qualified becoming in the node listing of backup in representing a little, creating one does not also have to declare to become and backs up the subclass of representing a little;

If it oneself is that backup is represented a little that one or more node is all stated, the node that has limit priority so becomes backup to be represented a little, determines that by other attributes relatively final backup represents a little if priority equates, flow process finishes;

Selection has the node of limit priority to be represented a little for backup, and relatively the definite final backup of other attributes is represented a little if priority equates.

Above-mentioned other attributes can include but not limited to node identification and IP address.

Above-mentioned optical-fiber network includes but not limited to ASON.

Above-mentioned inter-domain protocol includes but not limited to inter-domain routing protocol.

Main with representing a little or back up when representing point failure when occurring respectively in the same area not, in same area not, carry out active and standby representing respectively independently and a little switch protection.

Pass through technique scheme; the present invention can realize the effective protection of SPEAKER in the DDRP agreement; switch fast; the SPEAKER node failure can effectively reduce because may cause the business of entire domain outage probability to occur separately; simultaneously with standard DDRP protocol-compliant; algorithm complex is little, has succinct, reliable advantage.

Description of drawings

Accompanying drawing described herein is used to provide further understanding of the present invention, constitutes the application's a part, and illustrative examples of the present invention and explanation thereof are used to explain the present invention, do not constitute improper qualification of the present invention.In the accompanying drawings:

Fig. 1 is used for the flow chart that the optical-fiber network inter-domain protocol is represented guard method a little according to of the present invention;

Fig. 2 is the network structure's schematic diagram according to the embodiment of the invention;

Fig. 3 is according to the DDRP SPEAKER node schematic diagram in the hierarchical network of the embodiment of the invention;

Fig. 4 is the network condition schematic diagram when breaking down as SPEAKER and backup SPEAKER according to the embodiment of the invention;

Fig. 5 switches later network condition schematic diagram according to executing of the embodiment of the invention;

Fig. 6 is the renewal process schematic diagram of the information interaction between the active and standby SPEAKER in the territory according to the embodiment of the invention;

Fig. 7 is the main motion flow schematic diagram of carrying out with the SPEAKER node according to the embodiment of the invention;

Fig. 8 is after the backup SPEAKER according to the embodiment of the invention breaks down, the main motion flow schematic diagram of carrying out with the SPEAKER node; And

Fig. 9 is the whole technical proposal workflow diagram according to the embodiment of the invention.

Embodiment

Describe the present invention below with reference to the accompanying drawings in detail.

With reference to Fig. 1, be used for the guard method that the optical-fiber network inter-domain protocol represents a little and may further comprise the steps according to of the present invention:

Step a, the candidate qualification of disposing each network element, include qualification become main with represent point, qualified become to back up represent point or do not have qualification;

Step b, between node and node, set up after two-way communication is connected, each node is listed whole main using that may become respectively and is represented a little and back up the tabulation of representing network element a little, each network element is according to the candidate qualification then, declare oneself to be main represent point or do not make a speech with representing point or backing up, if it oneself is main with representing a little or backing up and represent a little not having the network element statement, then change step e over to;

Step c, all qualified become main with the tabulation of the network element in representing a little in, creating one also not have to declare to become the master and uses the subclass of representing a little;

Steps d oneself is main with representing a little if one or more network element is all stated, the network element that has limit priority so becomes the master with representing a little, determines that by comparing other attributes final master with representing a little, proceeds to step f if priority equates;

Step e, the network element that selection has limit priority is main with representing a little, relatively other attributes are determined finally main with representing a little if priority equates;

Step f, whole qualified becoming in the network element tabulation of backup in representing a little, creating one does not also have to declare to become and backs up the subclass of representing a little;

Step g oneself is that backup is represented a little if one or more network element is all stated, the network element that has limit priority so becomes backup to be represented a little, determines that by other attributes relatively final backup represents a little if priority equates, flow process finishes;

Step h, selection has the network element of limit priority to be represented a little for backup, relatively the definite final backup of other attributes is represented a little if priority equates.

Elect new backup to represent a little and can may further comprise the steps:

Whole qualified becoming in the network element tabulation of backup in representing a little, creating one does not also have to declare to become and backs up the subclass of representing a little;

If it oneself is that backup is represented a little that one or more network element is all stated, the network element that has limit priority so becomes backup to be represented a little, determines that by other attributes relatively final backup represents a little if priority equates, flow process finishes;

Selection has the network element of limit priority to be represented a little for backup, and relatively the definite final backup of other attributes is represented a little if priority equates.

Above-mentioned optical-fiber network includes but not limited to ASON.

The present invention relates to field of optical networks, be specifically related to represent in a kind of inter-domain protocol of ASON network host-slave protection method a little, also be applicable to that the inter-domain protocol of data network represents point protection simultaneously.

Fig. 2 has illustrated the situation and the network configuration of network node, is two ASON network domains and node situation, and the numeral in the node identification is the ID numbering, and the ID of other accompanying drawings also is like this.

The active and standby SPEAKER node election of Fig. 3 is finished, and square is SPEAKER, and two rhombuses are spare spes AKER.

The inner N11 of CD1 among Fig. 4 breaks down, and causes the main SPEAKER can't operate as normal, and the inner N22 of CD2 breaks down, and causing backing up SPEAKER can't operate as normal.The production of these two kinds of faults of CD1 and CD2 is fully independently, does not influence each other.

SPEAKER main-apparatus protection technical scheme of the present invention mainly is made of following several parts, comprises that respectively the protection of election, the information interaction between the active and standby SPEAKER of SPEAKER in the territory, active and standby SPEAKER is switched and new backup SPEAKER election.

Further, technical scheme of the present invention comprises following technical step:

The first step, the DDRP agreement starts, and finishes when main SPEAKER elects in the territory, finishes the election of backup SPEAKER.

In second step, the routing iinformation that active and standby SPEAKER carries out the place routed domain respectively gathers, and the interlayer communication function of corresponding routed domain (for example, feed up and feed down).

In the 3rd step, establish a communications link between the active and standby SPEAKER.Active and standby SPEAKER carries out routing iinformation critical data and state backup by this communication contact each other.

In the 4th step, active and standby SPEAKER detects survival condition separately mutually by communication contact.If testing result shows that main SPEAKER breaks down, then changed for the 5th step over to; Be equipped with SPEAKER and break down, then changed for the 6th step over to; Otherwise, continue to carry out this step.

In the 5th step, after backup SPEAKER detected main SPEAKER fault, backup SPEAKER started and carries out the correlation function of main SPEAKER.Be equipped with SPEAKER and become new main SPEAKER.

In the 6th step,, in this territory, re-elect SPEAKER fully according to the method for the first step.Flow process changes the 4th step in the technical scheme over to.

Especially, further comprise following technical step about main SPEAKER with the electoral package that is equipped with SPEAKER in the first step in the technical scheme:

Step 1, each node electrifying startup, the DDRP agreement on the node starts, and by the candidate qualification that webmaster or other means dispose each ASON network element, including qualification becomes main SPEAKER, and qualified becoming backed up SPEAKER or do not had qualification.

Step 2, setting up between node and the node after two-way communication connects, each node is listed the tabulation of whole ASON network elements that may become SPEAKER and backup SPEAKER respectively, each ASON network element is according to the candidate qualification situation of step 1 configuration then, declares oneself to be main SPEAKER or to be equipped with SPEAKER or not make a speech.If not have ASON network element statement oneself be main or be equipped with SPEAKER, then change step 5 over to, otherwise change next step over to.

Step 3, in whole qualified ASON network elements tabulations that becomes among the main SPEAKER, creates a subclass that does not also have to declare to become SPEAKER.

If one or more ASON network element of step 4 all states it oneself is main SPEAKER, the ASON network element that has limit priority so becomes main SPEAKER network element, determines final main SPEAKER network element if priority equates by comparing other attributes.Flow process finishes.

The ASON network element that step 5, selection have limit priority is SPEAKER, if priority equates then determines final SPEAKER network element according to other attributes of the described comparison of step 4.Flow process finishes.

Especially, described other attributes of step 4 comprise the sign of ASON node, relevant parameter such as IP address or other.Can be used as a kind of strategy disposes in advance in step 1.

Especially, after the same method,, can finish the election of SPEAKER fully by step 3～5 of repeating above-mentioned technical step.

After main SPEAKER and backup SPEAKER election were finished, described according to second step in the technical scheme, main SPEAKER carried out the interlayer communication of information, and relevant routing iinformation is gathered.

Fig. 6 is seen in communication and critical data backup in the technical scheme between the described active and standby SPEAKER of the 3rd step.Fig. 6 has illustrated in the territory between the active and standby SPEAKER process of query State mutually, and the operating process of switching when speaker breaks down in backup, simultaneously, and the information interaction renewal process in Fig. 6 representative domain between the active and standby SPEAKER.

In the technical scheme the 4th step comprises mutual detection and the corresponding handling process between the active and standby SPEAKER.Especially, in the technical scheme the 4th step about main SPEAKER be equipped with SPEAKER to detect the concrete steps that survival condition scheme separately comprises mutually as follows:

Step 1, the main time interval of sending STATUS ENQUIRY message with the responsible tracking detection of SPEAKER startup specific timing device (being assumed to be timer 1).

Step 2, carry out synchronously with step 1, whether spare spes AKER starts timer (being assumed to be timer 2) and is responsible for detecting and has inquiry or order to send with SPEAKER from main in the stipulated time in the step 2.In this position, if the timer 1 of main usefulness SPEAKER is at first then, then enter step 3, if the timer 2 of backup SPEAKER is at first then, then enter step 6.

Step 3, according to the timer of step 1, if at first the master with the SPEAKER timer time then, then the master sends out message with SPEAKER and gives backup SPEAKER, the state of inquiry backup SPEAKER, simultaneously, mainly start a timer (timer 3) with SPEAKER and be responsible for inquiry whether to back up response time of SPEAKER overtime.While pause timer 1.

Step 4, according to step 3, if backup SPEAKER receives main Query Information with SPEAKER, then back up SPEAKER and send response, the timer 2 that resets has simultaneously shown that inside receives query message at the appointed time.

Step 5, according to step 4, timer 3 then before, main receive the response message of backup SPEAKER with SPEAKER, then main with SPEAKER timeing closing device 3, be illustrated in the stipulated time to receive response message, if it is normal that while message represents to back up the SPEAKER state, then main with SPEAKER replacement timer 1, all are normal to show backup SPEAKER, in this case, again return step 1, if then at timer 3.Do not receive that also response or the response that receives represent to back up the SPEAKER mistake, the main timing that does not recover timer 1 with SPEAKER this moment.According to the step of second in the technical scheme, then need to enter the 6th step of technical scheme.The detailed process of this part is seen Fig. 8.

If backup SPEAKER timer then in step 6, the step 2.Then under the timer 2 situation then, backup SPEAKER initiates a message and inquires about main state with SPEAKER, and pause timer 2 starts timer (being assumed to be timer 4) simultaneously and monitors with the response of SPEAKER main.

Step 7, according to step 6, if timer 4 then before, backup SPEAKER receives main response message with SPEAKER, and message shows that main all are normal with SPEAKER, then backs up SPEAKER timeing closing device 4, recovers and restarts timer 2.This moment, the recovering state of whole proposal arrived step 1, otherwise entered the 5th step of technical scheme.Idiographic flow is seen Fig. 7.Fig. 7 represents the main motion flow with the SPEAKER fault of spare spes AKER node discovery.

Especially, comprise following technical step according to the step of the 5th in the technical scheme:

Step 1, backup SPEAKER at first check the database of the main SPEAKER of local backup.

Step 2, according to the result of step 1, if the database in the backup SPEAKER is effective, then backs up SPEAKER and can send statement to the node that exists and oneself be the message of SPEAKER.And wait for the answer of other nodes.Change the processing of step 3 over to.If invalid, then return in the first step in the technical scheme, carry out the election of SPEAKER again.

Step 3, receive the answer of other network nodes according to the result of step 2 after, the state that backup SPEAKER can change itself, the state of oneself is set to main SPEAKER.

After step 4, step 3 were finished, this main SPEAKER other nodes in network sent updating message.Wait for database update.

After step 5, step 4 were finished, this main SPEAKER can send backup SPEAKER node election message, backs up the election of SPEAKER again,

Step 6, see the detailed step of the backup SPEAKER election in the first step in the technical scheme.

Above step is seen Fig. 9.

Embodiment 1:

In conjunction with Fig. 2 and Fig. 3, be that example elects the specific embodiments of active and standby SPEAKER to be elaborated to the first step in the technical scheme to elect main SPEAKER:

The first step is set up two territory CD1 and CD2 according to the step 1 of the technical scheme first step, and network diagram is seen Fig. 2.Each network element powers on then, and each network element among CD1 and the CD2 all claims oneself to be main SPEAKER.

Second step, set up two-way connection by sending packet between the node, each node is all set up database, and comprising all being claimed as in the database oneself is main SPEAKER node, and according to shown in Figure 2, all node all claims oneself to be main SPEAKER.

The 3rd step, create a subclass that does not also have declaration to become SPEAKER, this subclass is empty.

In the 4th step, in this step, suppose the limit priority that has of node ID number minimum, so node N11 and the node N21 among the CD2 in the CD1 become the main SPEAKER in territory separately respectively.

According to same quadrat method, elect N12 and N22 and be the backup SPEAKER of inside, territory separately.

So far, the election of SPEAKER node is finished in the network.The figure that election is finished sees Fig. 3.

Embodiment 2:

In conjunction with Fig. 3 and Fig. 6, having elected main SPEAKER N11 with embodiment 1 territory CD1 is example to the detection step of backup SPEAKER N12, active and standby SPEAKER in the technical scheme is detected mutually be specifically described:

The first step, the N11 network element starts timer 1, and the N12 network element starts timer 2.

Second the step, N11 timer 1 then, pause timer 1 changes next step over to.

In the 3rd step, N11 network element transmit status query messages is given N12.The N11 network element starts timer 3 simultaneously.

The 4th step, timer 3 then before, the N11 network element receives N12 and responds successfully N11 network element timeing closing device 3.

In the 5th step, N11 network element reset timer 1 carries out the next round timing again.So far main step with SPEAKER detection backup SPEAKER is finished.Change next step over to.

The 6th step, continue, N12 network element timer 2 then, pause timer 2, inquiry N11 state, and start timer 4.

The 7th the step, timer 4 then before, the N12 network element receives the normal status response message of N11 network element, timeing closing device 4.

The 8th step, N11 network element reset timer 2, reclocking.So far, the backup SPEAKER step that detects main SPEAKER node is finished.

So far, the main detection with SPEAKER backed up SPEAKER and the main process end with SPEAKER of backup SPEAKER detection.

Embodiment 3:

In conjunction with Fig. 3, Fig. 4 and Fig. 8, having elected main SPEAKERN21 and backup SPEAKER N22 with territory CD2 among the embodiment 1 is example, and the switching action that is equipped with the initiation of SPEAKER fault in the technical scheme is specifically described, execution in step is as follows:

The first step, the people is the power supply of closing N22, sees the territory CD2 among Fig. 4.

Second step, the timer of N21 then after, inquiry N22 state because the N22 fault, therefore inquiry failure.

The 3rd step, the message of re-electing backup SPEAKER that N21 sends.Re-elect backup SPEAKER.

In the 4th step, the CD2 inner correlation step of carrying out in the embodiment 1 in territory is re-elected out backup SPEAKER N23.

The 5th step, data sync between N22 and the N23, SPEAKER database synchronously.

So far, the switching action that causes of backup SPEAKER fault is finished.Referring to the territory CD2 among Fig. 5.

Embodiment 4:

In conjunction with Fig. 3 and Fig. 7, having elected main SPEAKER N11 and backup SPEAKER N12 with embodiment 1 territory CD1 is example, and the main switching action that causes with the SPEAKER fault in the technical scheme is specifically described:

The first step, the people is the power supply of closing N11.

Second step, N22 timer 3 then after because the N11 fault, N21 status poll failure.

In the 3rd step, N22 checks backup database, owing to carried out the operation of SPEAKER database synchronization originally, so the data base querying success.

In the 4th step, it oneself is the statement of main SPEAKER that N22 sends to other nodes of inside, territory.

In the 5th step, N22 receives the response of each interior nodes of territory.

In the 6th step, N22 change state is the main SPEAKER that uses.

In the 7th step, N22 initiates election backup SPEAKER, the inner election backup SPEAKERN23 in the territory, and network diagram is seen the territory CD1 among Fig. 5.

In the 8th step, carry out data sync between N22 and the N23.

In the 9th step, N22 initiates network information update operation in the territory.

So far, the masterslave switchover action that causes of main SPEAKER mistake is finished.

Embodiment 5:

In conjunction with Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7 and Fig. 8, having elected main SPEAKER N11 with embodiment 1 territory CD1 is example to the detection step of backup SPEAKER N12, is specifically described occurring the switching action that active and standby SPEAKER fault causes in the same area not respectively:

The first step as Fig. 2, is set up two territory CD1 and CD2, and each network element electrifying startup changes next step in the territory.

Second step, carry out SPEAKER election in the territory, SPEAKER N11 and backup SPEAKER N12 are finished in CD1 election in territory, and SPEAKER N21 and backup SPEAKERN22, election results such as Fig. 3 are finished in election in the CD2 of territory.

In the 3rd step, the data sync among territory CD1 and the territory CD2 between the active and standby SPEAKER is finished.

In the 3rd step, the people breaks down N11 for making among the CD1 of territory, and the backup N22 among the CD2 of territory breaks down.Schematic diagram is seen Fig. 4.

The 4th step, the timer of SPEAKER then in each territory, therefore territory CD1 backup SPEAKERN12 inquiry N11 state failure, territory CD2 master uses the failure of SPEAKR N21 inquiry backup SPEAKER N22 state, so independently carries out the masterslave switchover action in territory CD1 and the CD2 respectively.Concrete steps are seen embodiment 3 and embodiment 4.

In the 5th step, the active and standby SPEAKER in each territory switches and finishes, and the SPEAEKR in the CD1 of territory is N12, and the main SPEAEKR in the CD2 of territory is N22.

In the 6th step, SPEAKER N12 each territory in and N22 initiate to re-elect backup SPEAKER respectively, and N13 is backup SPEAKER in the election territory CD1, and N23 finishes later network diagram and sees Fig. 5 for backing up SPEAKER, switching in the CD2 of territory.

In the 6th step, the active and standby SPEAKER in each territory carries out data sync and renewal.

So far, not the active and standby SPEAKER in the same area respectively switching action finish.

According to above embodiment as can be known, the protection mechanism that the technical program adopts active and standby SPEAKER to back up mutually possesses fast quick-recovery SPEAKER function, switches the high advantage of efficient.

The above is the preferred embodiments of the present invention only, is not limited to the present invention, and for a person skilled in the art, the present invention can have various changes and variation.Within the spirit and principles in the present invention all, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims

1. one kind is used for the optical-fiber network inter-domain protocol and represents a little guard method, it is characterized in that, may further comprise the steps:

Step S102, main with representing a little and backup is represented a little in the election territory, described master uses and represents a little and nodal information and the routing iinformation that a little converges the territory, place respectively represented in described backup, and the execution interlayer communication;

Step S104, described main with representing a little and described backup establishes a communications link between representing a little, be used to carry out that the routing iinformation critical data backs up and state backs up;

Step S106, described main with represent a little and described backup represent a little between by the described mutual detection survival condition separately that communicates to connect, if described backup is represented a little to detect describedly main with representing fault appears a little, then described backup is represented a little becomes new master with representing a little, elect new backup to represent a little, be back to step S104, if describedly main a little detect described backup and represent and fault a little occurs with representing, then elect new backup to represent a little, be back to step S104, do not break down if all detect, then repeated execution of steps S106.

2. guard method according to claim 1 is characterized in that, in step S102, elects described main with representing a little and described backup is represented a little and be may further comprise the steps:

Step b, between node and node, set up after two-way communication is connected, each node is listed whole main using that may become respectively and is represented a little and back up the tabulation of representing node a little, each node is according to described candidate qualification then, declare oneself to be main represent point or do not make a speech with representing point or backing up, if it oneself is main with representing a little or backing up and represent a little not having the node statement, then change step e over to;

Step e, the node that selection has limit priority is main with representing a little, more described other attributes determine that final master is with representing a little if priority equates;

Step g oneself is that backup is represented a little if one or more node is all stated, the node that has limit priority so becomes backup to be represented a little, determines that by more described other attributes final backup represents some the flow process end if priority equates;

Step h, the node of selecting to have limit priority is represented a little for backup, the definite final backup of more described other attributes is represented a little if priority equates.

3. guard method according to claim 2 is characterized in that, described other attributes comprise node identification and IP address.

4. guard method according to claim 1 is characterized in that, in step S106, described main with represent a little represent a little with described backup between mutually detection survival condition separately may further comprise the steps:

Step P1, described main with representing a startup first timer, be used to follow the tracks of and detect the time interval of sending STATUS ENQUIRY message, a startup second timer is represented in described backup, and whether be used to detect has inquiry or order to send with representing a little from described master in the stipulated time;

Step P2 if described first timer at first then, then enters step P3, if described second timer at first then, then enters step P6;

Step P3 describedly main a little send out message and represents a little to described backup with representing, and inquires about described backup and represents a little state, and start the 3rd timer, and it is overtime to be used to inquire about the response time whether described backup is represented a little, and suspends described first timer;

Step P4 a little receives the described main Query Information of representing a little of using if described backup is represented, and then described backup is represented and a little sent response, and the described second timer that resets has simultaneously shown that inside receives query message at the appointed time;

Step P5, if described the 3rd timer then before, describedly main a little receive described backup and represent a little response message with representing, then described master a little closes described the 3rd timer with representing, being illustrated in the stipulated time receives response message, if and message to represent that dotted state is represented in described backup normal, then described main with representing a described first timer of replacement, show that described backup represent point all are normal, be back to step P1, if also do not receive then that at described the 3rd timer response or the response that receives represent that described backup represents a mistake, then described master does not a little recover described first timer with representing, judge described master and fault a little occurs with representing, suspend and detect, execution is switched;

Step P6, described backup represents that the inquiry that a little initiates a message is described main to suspend described second timer with the state of representing a little, and starts the 4th timer and come mainly to use the response of representing a little to monitor to described;

Step P7, if described the 4th timer then before, described backup is represented and is a little received the described main response message of representing a little of using, and message shows describedly mainly puts with representing that all are normal, and then described backup is represented and a little closed described the 4th timer, recovers and restarts described second timer, be back to step P1, represent and fault a little occurs otherwise judge described backup, suspend and detect, execution is switched.

5. guard method according to claim 1 is characterized in that, in step S106, described backup is represented a little becomes new master with representing a little, elects new backup to represent a little and may further comprise the steps:

Step T1, described backup is represented and is a little at first checked the main with the database of representing a little of local backup;

Step T2, effective if a little interior database is represented in described backup, then backup is represented a little to send to the node that exists and is stated that oneself is the message of representing a little, and waits for the answer of other nodes, if invalid, then proceeds to step S102;

Step T3 receives after the answer of other network nodes, and the state of a change itself is represented in described backup, and the state of oneself is set to new master with representing a little;

Step T4, the node of other in network sends updating message to described new master with representing a little, waits for database update;

Step T5, described new master represent a little to send to back up and represent a node election message, back up the election of representing a little again.

6. guard method according to claim 1 or 5 is characterized in that, elects new backup to represent a little and may further comprise the steps:

Selection has the node of limit priority to be represented a little for backup, and more described other attributes determine that finally backup is represented a little if priority equates.

7. guard method according to claim 6 is characterized in that, described other attributes comprise node identification and IP address.

8. according to each described guard method in the claim 1 to 5, it is characterized in that described optical-fiber network comprises ASON.

9. according to each described guard method in the claim 1 to 5, it is characterized in that described inter-domain protocol comprises inter-domain routing protocol.

10. according to each described guard method in the claim 1 to 5, it is characterized in that, main when occurring respectively in the same area not with representing a little or back up when representing point failure, in described not same area, carry out active and standby representing respectively independently and a little switch protection.